BEST-MSU
(2021)Objective
To compare outcomes of acute ischemic stroke patients managed by mobile stroke units (MSUs) versus standard emergency medical services (EMS).
Study Summary
• MSUs led to faster tPA delivery and improved functional outcomes at 90 days.
• More patients in MSU group received tPA and within the 'golden hour'.
• More patients in MSU group received tPA and within the 'golden hour'.
Intervention
Alternating-week assignment to MSU or EMS care. MSUs were equipped with CT scanners and provided pre-hospital tPA if eligible.
Inclusion Criteria
Suspected acute stroke within 4.5 hours of onset, eligible for thrombolysis.
Study Design
Arms: MSU vs. EMS
Patients per Arm: MSU: 886 enrolled (617 eligible for tPA), EMS: 629 enrolled (430 eligible for tPA)
Outcome
• mRS 0–1 at 90 days: 53.5% (MSU) vs. 45.5% (EMS); OR 2.14; P<0.001.
• Median onset-to-tPA time: 72 min (MSU) vs. 108 min (EMS).
• Mortality: 8.9% (MSU) vs. 11.9% (EMS).
• Median onset-to-tPA time: 72 min (MSU) vs. 108 min (EMS).
• Mortality: 8.9% (MSU) vs. 11.9% (EMS).
Bottom Line
Mobile stroke unit care was associated with a significantly greater likelihood of achieving functional independence (mRS 0–1) at 90 days compared with standard EMS transport.
Major Points
- First multicenter RCT demonstrating that mobile stroke units (MSUs) improve functional outcomes: 55.0% vs 48.1% achieved excellent outcomes (mRS 0–1) at 90 days (OR 1.30, 95% CI 1.02–1.66, P=0.04).
- Cluster-randomized design across 7 US sites — MSU availability (on-day vs off-day) determined allocation, minimizing selection bias while reflecting real-world deployment.
- Median onset-to-treatment time reduced by 36 minutes (72 vs 108 min, P<0.001) — a massive time savings translating directly to better outcomes per the 'time is brain' principle.
- MSU provided prehospital CT imaging, teleneurology consultation, and IV tPA initiation in the field — a complete acute stroke evaluation before hospital arrival.
- No increase in symptomatic ICH (2.6% MSU vs 3.0% EMS) or serious adverse events — safety comparable to standard EMS despite prehospital thrombolysis.
- Mortality trend favored MSU (10.4% vs 12.6%) though not statistically significant — the benefit was primarily in functional recovery rather than survival.
- NNT of approximately 14 for one additional patient achieving excellent outcome — compelling for a systems-level intervention.
- Diverse patient population (only 40–43% White) — one of the most racially diverse stroke trials, addressing health equity concerns.
- Cost-effectiveness remains a major barrier — MSUs cost $1–2M annually to operate, and BEST-MSU does not address whether the functional improvement justifies the healthcare system investment.
- Established the evidence base that led to expanded MSU programs across major US cities and influenced the 2024 AHA/ASA guidelines recommending MSU deployment in high-volume areas.
Study Design
- Study Type
- Prospective, multicenter, cluster-randomized controlled trial
- Randomization
- Yes
- Blinding
- Outcome assessors blinded
- Sample Size
- 617
- Follow-up
- 90 days
- Centers
- 7
- Countries
- United States
Primary Outcome
Definition: Proportion of patients with modified Rankin Scale (mRS) score of 0–1 at 90 days among those with confirmed ischemic stroke
| Control | Intervention | HR/OR | P-value |
|---|---|---|---|
| 48.1% | 55.0% | - (1.02–1.66) | 0.04 |
Limitations & Criticisms
- Cluster-randomized design (MSU on-day vs off-day) may allow unmeasured confounding — patients presenting on MSU days may differ from off-days in ways not captured by baseline characteristics.
- Conducted exclusively in well-resourced US urban areas — generalizability to rural settings, lower-volume centers, or resource-limited healthcare systems is unknown.
- MSU operating cost ($1–2M/year per unit) was not addressed — cost-effectiveness analysis is critical for policy decisions about MSU deployment.
- The primary outcome was mRS 0–1 (excellent outcome) rather than mRS 0–2 (functional independence) — the mRS 0–2 endpoint did not reach significance (P=0.27), suggesting the benefit is concentrated at the top of the functional spectrum.
- Enrollment spanned 6 years (2014–2020) — practice changes during this period (thrombectomy expansion, TNK adoption) may create period effects that confound interpretation.
- COVID-19 pandemic overlap (2020) may have affected final enrollment months, EMS response patterns, and hospital protocols.
- Cannot separate the effect of faster treatment from the effect of improved triage (routing LVO patients directly to thrombectomy centers) — both contribute to MSU benefit.
- Long-term outcomes beyond 90 days were not assessed — durability of functional gains and downstream healthcare utilization unknown.
- Stroke mimics receiving tPA in the prehospital setting (inherent to the MSU model) were excluded from the efficacy analysis — real-world MSU mimic rates and consequences not fully characterized.
Citation
Grotta JC, Yamal SL, Parker SA, et al. Effect of a Mobile Stroke Unit on Functional Outcomes Among Patients With Ischemic Stroke: The BEST-MSU Randomized Clinical Trial. JAMA. 2021;325(5):504–513.